摘要
网格计算是近几年来出现的一种新兴技术。网格(Grid)的概念产生于上世纪90年代中期,是从电力网的概念借鉴而来的。网格的目的是利用互联网把分散在不同地理位置的电脑组织成一台虚拟的超级计算机,实现计算资源、存储资源、数据资源、信息资源、软件资源、通信资源、知识资源、专家资源的全面共享。
网格的概念被提出后,网格技术一直在不断发展。其中,网格体系结构对网格技术的发展起着十分重要的作用。网格体系结构是关于如何建造网格的技术,是网格的骨架和灵魂,是网格最核心的技术。只有建立合理的网格体系结构,才能够设计和建造好网格,才能够使网格更有效地发挥作用。随着互联网的发展,在以IBM为代表的工业界的影响下,考虑到Web技术的影响后,全球网格论坛结合Web服务技术提出了开放网格服务体系结构OGSA。OGSA最突出的思想是以“服务”为中心,并且提出了一种以服务为中心的模型。开放网格服务体系结构的出现使得网格服务与网络服务日趋融合,网格体系结构总体上正走向标准化。
网格的出现使得那些要求大量计算的科学研究、庞大的数据处理甚至是商业操作等各种应用都能够更快更完美地实现。但是这些应用的底层结构都比较复杂并且难以管理,于是各种各样的服务与门户被开发出来,在这种情况下网格Portal应运而生。网格Portal是网格系统的重要组成部分,是连接网格用户和网格系统的桥梁。通过网格Portal,网格用户可以很方便地访问网格资源和服务,执行和监控网格应用,实现对网格的各种操作。网格Portal为用户提供了基于浏览器的统一访问界面,降低了复杂网格系统的使用难度,推动了各种网格应用的快速发展。
本文以开放网格体系结构OGSA为基础,搭建了网格门户的开发环境,基于此开发环境,研究并实现了基于Hibernate和面向服务两种模式的简单的网格Portal。进一步研究了网格环境中基于GridFTP的数据传输机制,实现了以Java CoG为基础并使用GridFTP协议的数据传输;并结合FTP和GridFTP协议的优势,针对教育资源网格自身的特点,提出了复合型传输模式ERFTP,这也是本文的创新所在。
在前人工作的基础上,本文具体的研究和创新工作如下:
(1)本文研究了网格的体系结构、功能特征以及网格Portal的相关技术,重点研究了开放网格体系结构OGSA,详细比较了各种网格门户开发工具,并选择合适的网格门户开发工具搭建了网格门户环境。
(2)本文详细介绍了Hibernate技术,研究了利用Hibernate技术进行网格Portal设计的相关内容,并在网格Portal的开发中引入了Hibernate技术,方便了软件开发中对数据库的操作。通过利用Hibernate技术进行教育资源网格Portal的开发,充分表明Hibernate技术的使用大大提高了软件项目的开发效率。
(3)本文在开放网格服务体系结构的基础上,阐述了网格服务的编写及部署过程,并把网格服务整合到网格Portal中,极大地方便了用户对于网格资源的使用和操作。
(4)本文介绍了网格环境下基于Java CoG的数据传输机制,实现了网格环境下利用GridFTP协议进行数据传输的功能,并结合FTP和GridFTP的优势,提出了复合型传输模式ERFTP,并把该研究应用到国家自然科学基金项目《基于因特网的教育资源网格体系结构和服务理论研究》的副本管理模块中,运行效果良好。这也是本文的创新点所在。
Grid computing is a new technology emerging in recent years. The grid concept stems from the mid-1990s and it comes from electric network. The Grid originally makes use of computers in different places and combines them through Internet to form a super virtual computer to help realize the comprehensive sharing of computing resources, storage resources, data resources, information resources, software resources, communication resources, knowledge and expert resources and so on.
Since the grid concept was proposed, the grid technology has developed rapidly, and grid architecture has played an important role in the development of the grid technology. Grid architecture focuses on how to build a grid and it is the backbone and core technology of the grid. It’s important to realize that only on the basis of a reasonable grid architecture can we design a suitable grid and make full use of it. With the development of the Internet and under the influence of industry leaders like IBM, the Global Grid Forum, after taking into account the impact of Web technology and absorbing the Web services technology, proposed the Open Grid Services Architecture (OGSA). The most prominent feature of the OGSA is focusing on“service,”and it has proposed a service-oriented model. The OGSA combines Grid service with Web service, and the Grid architecture in general is taking the path of standardization.
The emergence of the grid makes possible a wide range of applications, such as scientific research that requires a large amount of calculation, large data-processing and even the commercial operation, which can now be done more quickly and more satisfactorily. But the underlying structure of these applications is more complex and difficult to manage, so all kinds of services and portals have been developed. It is under such circumstances that grid portal has come into existence. Grid portal is one of the most important parts of a grid. It is the bridge that links users and the grid, through which users can visit various grid resources and complete different operations. Grid portal provides a uniform access interface based on web browser, reducing the complexity of using the grid and thus pushing forward the grid development.
Firstly this paper sets up a grid portal development environment on the basis of the OGSA. Then based on this development environment, it studies and realizes the development of two models of Hibernate-based and service-oriented simple grid portal. It goes further to study the data transfer mechanism based on GridFTP under the grid environment and realizes data transmission using GridFTP agreement based on Java CoG. Finally, by combining the advantages of FTP and GridFTP, this paper has proposed a complex transfer protocol, ERFTP, in view of the characteristic of the Educational Resource Grid. This is where the innovation of this paper lies.
On the basis of previous researches, the concrete research work and innovation of the paper can be described as follows:
(1)The paper studies the architecture, the function characteristic of the grid as well as the related technologies of grid portal, and in particular the OGSA. Having made a detailed comparison of the different grid portal development kits, it selects the appropriate grid portal development tools to build a grid portal environment.
(2)The paper introduces the Hibernate technology in detail and explores the technology of developing the Grid Portal using Hibernate. It facilitates the operation of the database, through the introduction of Hibernate technology in the Portal development. With the development of a practical Educational Resource Grid, it fully shows that the use of the Hibernate technology has greatly improved the efficiency of the software development projects.
(3)Based on the Open Grid Services Architecture, the paper elaborates on how to write and deploy a Grid Service in detail. Meanwhile, it explains how to combine the Grid Service with the grid portal, which will provide great convenience for the users in their use and management of the grid resources.
(4)The paper introduces the data transfer mechanism based on Java CoG under the grid environment and realizes data transmission using GridFTP agreement. Combining the advantages of FTP and GridFTP, this paper has proposed a composite transfer protocol ERFTP. This part of the research has been applied in the replica management module of the National Natural Science Foundation of China project“The Architecture and Service Theory of Educational Resource Grid Research Based on Internet”, and the effect turns out to be very good. This is the key innovation presented by this paper.
引文
[1]都志辉,陈渝,刘鹏.网格计算[M].北京:清华大学出版社.2002.
[2]I.Foster, H.Kishimoto, A.Savva, D.Berry, et al.The Open Grid Services Architecture [J]. Version1.5.24 July 2006.
[3]I.Foster,Carl Kesselman.The Grid Blueprint for a New Computing Infrastructure.China Machine Press.2005. [ 4 ] National Grid. The statement of the use of system charging methodology [EB/OL]. http: //www.nationalgrid.com/uk. 2005.
[5]苏振强,徐佳维,蔡文生,邵学广.基于Globus Toolkit和SunTM ONE Grid Engine构建计算网格[J].计算机与应用化学.2005,22(2):81~84.
[6]IBM网格计算进入业务应用.计算机教育.2003(1):48~49.
[7]Astrophysics Simulation Collabrotary Portal.http://www.ascportal.com/.
[8]林楠.基于双层中间件的网格数据库访问与集成[J].微计算机信息.2009,25(1-3):182~183,208.
[9]Ian Foster,Carl Kesselman,Steven Tuecke. The Anatomy of Grid:Enabling Scalable Virtual Organizations, International Journal of Supercomputer Applications, 15(3),Sage Publications,2001,USA.
[10]中国教育科研网格公共支撑平台工作组.中国教育科研网格公共支撑平台设计规范[M].北京:清华大学出版社.2004.
[11]徐志伟,冯百明,李伟.网格计算技术[M].北京:电子工业出版社.2004.
[12]都志辉,陈渝,刘鹏等.以服务为中心的网格体系结构OGSA[J].计算机科学.2003,30(7):26~29.
[13]Karl Czajkowski, Donald F Ferguson, Ian Foster et al. The WS-Resource Framework. Mar 05 2004.
[14]The World Wide Web Consortium.Web Services Activity. http://www.w3.org/2002/ws/.Jan 24 2008.
[15]I. Foster, D.Gannon, H. Kishimoto, Jeffrin J.Von Reich.Open Grid Services Architecture Use Cases. Oct. 28, 2004.
[16]朱民,张德贤.网格体系结构综述[J].网络安全技术与应用.2009(2):69~71.
[17]Globus Alliance. Globus Toolkit4. http://www.globus.org/toolkit/.Jan 3 2008.
[18]Ian Foster, Carl Kesselman, Jeffrey M. Nick, Steven Tuecke. The Physiology of the Grid: An Open Grid Services Architecture for Distributed Systems Integration, June 22 2002.
[19]Christian Wege.Portal Server Technology.IEEE Internet Computing,Vol. 6,No.3,7377, 2002.
[20]F Bellas.Standards for Second-Generation Portals. IEEE Internet Computing, 2004,Vol.8, No.2,54~60.
[21]MaoZhen Li, Mark Baker.网格计算核心技术[M].北京:清华大学出版社.2006.
[22]G Andronico, R Barbera, A Falzone, et al.The Genius Web portal: Grid Computing Made Easy. International Conference on Information Technology, 2003.
[23]JSR 168: Portlet Specification. Java Community Process. http://www.jcp.org/en/jsr/detail?id=168. Oct 10 2003.
[24]A Abdelnur, S Hepper. The Java Portlet Specification[OL], http://en.wikipedia.org/wiki/JSR_168. 2009.
[25]陈博,迟学斌等.计算网格Portal的设计与实现[J].微电子学与计算机.2004,21(9):15-18.
[26]D Gannon, G Fox, M Pierce et al.Boisseau, Grid Portals: A Scientist’s Access Point for Grid Services(DRAFT 1). Sep 19 2003.
[27]使用Grid Portal Toolkit 3构建网格门户. http://www.ibm.com/developerworks/cn/grid/ gr-gridport/index.html.2004.
[28]王璿.基于gridport的计算服务网格门户的设计与实现[J].燕山大学学报.2007,31(2):169~173.
[29]储凡静.基于GridSphere的服务网格门户技术的研究与应用的研究与实现[D].济南:山东师范大学硕士论文,2008.
[30]朱永虎,张选平,谭小鹏,刘君瑞.Jetspeed:企业信息门户的实现[J].计算机应用研究.2003,20(3):155~156,160.
[31]陈美荣,蔡洪斌,李武林,杨春.基于GridSphere的网格门户设计与实现[J].计算机应用研究.2006,23(2):213~215.
[32]孙卫琴.精通Hibernate:Java对象持久化技术详解[M].北京:电子工业出版社.2005.
[33]龙海阔.基于OGSA的网格Portal应用研究[D].杭州:浙江大学硕士论文,2007.
[34]袁满,张士迁,刘雪梅,郭玲玲,李瑞芳.基于网格的分布式异构平台数据传输机制探讨[J].大庆石油学院学报.2005,29(6):109~111,130.
[35]Globus Toolkit Tutorial. http://www.globusconsortium.org/ tutorial/. May 2006.
[36]IanFoster. GlobusToolkitVersion4:Software for Service-Oriented Systems. J.Comput. Sci. & Technol. July 2006,Vol.21,No.4,pp.513~520.
[37]Shan Jiu-long, Chen Guo-liang, He Jia-hua et al. Grid society: a system view of grid and P2P environment[C]. In: Proceedings of the International Workshop on Grid and Cooperative Computing GCC2002.
[38]刘文杰.基于Globus的网格应用关键技术研究[D].武汉:武汉大学硕士论文,2004.
[39]刘国恩.基于Globus工具包的网格计算系统研究[D].北京:北京航空航天大学硕士论文, 2004.
[40] The Globus Project.Globus Toolkit(TM) Developer Tutorial.Grid Security Infrastructure. 16 Nov 2005.
[41]李文中,周俊,顾铁成,李春洪,陆桑璐,陈道蓄.基于Globus的流媒体网格信息服务系统GT[J].小型微型计算机系统.2005,26(12):2074~2077.
[42]Steven Tuttle, Ami Ehlenberger, Ramakrishna Gorthi, Jay Leiserson, Richard Macbeth et al.International Technical Support Organization.Understanding LDAP Design and Implementation. June 2004.
[43]夏昕,曹晓钢,唐勇等著.深入浅出Hibernate[M].北京:电子工业出版社.2005.
[44]梁立新著.项目实践精解:基于Struts-Spring-Hibernate的Java应用开发[M].北京:电子工业出版社.2006.
[45]陶勇,李晓军著.Hibernate ORM最佳实践[M].北京:清华大学出版社.2007.
[46]冯征,杨寿保,陈华平.基于Portlet框架面向服务的网格Portal实现[J].计算机工程.2004 (7):91-92,115.
[47]关键.语义网格Portal的应用研究[D].南京:东南大学硕士论文,2006.
[48]GridFTP Universal Data Transfer for the Grid[EB/OL]. http://www.globus.org/toolkit/docs/3.0/ gridftp/C2WPdraft3.pdf. Sep 5 2000.
[49]王竑,黄红梅,杨文伟.基于JavaCoGKit的网格应用[J].计算机工程.2004,30(6):79~80,154.
[50]王永辉,蔡洪斌,杨春,周明天.网格环境中文件传输服务的研究与实现[J].计算机应用研究.2006(1):201~203.
[51]Karl Czajkowski, Don Ferguson, Ian Foster,et al. From Open Grid Services Infrastructure to WS-Resource Framework: Refactoring & Evolution,Version 1.1.Mar 5,2004.
[52]应宏,李梦蛟,刘福明,钟静,鄢沛.基于GridFTP的数据传输设计与实现[J].计算机科学.2006,33(5):147.
[53]Java GridFTP client-programmer guide [EB/OL]. http://www-unix.globus.org/cog/jftp/guide.html.
[54]China National Grid. http://www.cngrid.org/web/guest/home.
[55] Java CoG Kit Installation Guide. http://wiki.cogkit.org/index.php/Java_CoG_Kit_Installation_Guide. Aug 27 2006.
[56] GT 4.0 MyProxy: User's Guide.http://www.globus.org/toolkit/docs/4.0/security/myproxy/ user-index.html.
[57] The Globus Toolkit 4 Programmer’s Tutorial. Borja Sotomayor.2005.
[58] Global Information Grid. http://en.wikipedia.org/wiki/Global_Information_Grid. Dec 21 2008.
[59]黄理灿,刘元安.e-Science网格及其可扩展性体系结构[M].北京:北京邮电大学出版社.2005.